Changes in water quality resulting from human activities at the land surface in the Barnegat Bay watershed can be broadly characterized as resulting from either point sources or nonpoint sources of contamination. Point sources are commonly localized and originate from one specific source, such as wastewater discharges, leaky underground storage tanks, landfills, pipelines, and surface impoundments. Nonpoint sources of contamination do not result from activities in only one specific point. Rather, they are the cumulative effect of similar activities taking place at a number of locations within a given watershed, which together result in degradation of water quality within the watershed. Nonpoint sources of contamination can occur at widespread locations extending up to hundreds of square miles and over broad geographic areas. Nonpoint sources of contamination include activities occurring on agricultural, residential, and commercial properties to which fertilizers and pesticides have been applied, atmospheric inputs of airborne pollutants, animal feedlots, and right-of-ways (e.g., highway and railway borders) where deicing salts and pesticides are applied. On-site septic systems in residential areas, when in large numbers and closely spaced, also can be considered nonpoint sources of contamination (Vowinkel and Siwiec, 1991).

The aggregate effect on water resources of different human activities occurring in common land-use categories can be evaluated by examining the relations between observations of water quality of streams and ground water, associated constituent loadings, and the land-use composition of corresponding source areas. This approach has been used in many water resource investigations in southern New Jersey and surrounding areas.

The distribution of land use in 1994/95 within the Barnegat Bay watershed was determined to be the following: forested, 45.9%; wetlands (both tidal and freshwater), 25.2%; urban/residential, 19.5%; agricultural land/grasslands, 6.6%; barren lands, 1.9%; and water bodies (lakes, impoundments, reservoirs), 0.9%. These land-use percentages were determined in part by the use of a Geographic Information System and categories defined by Fegeas et al. (1983).

Impervious surfaces directly affect watersheds. The process of urbanization has a profound impact on the hydrology, morphology, water quality, and ecology of surface waters (Schueler, 1998). Impervious cover, which is greatest in urban areas, is a useful indicator with which to measure the impacts of land development on aquatic systems. It directly influences streams through the increase in surface runoff during storm events. Some of the influences of impervious cover on the annual water balance are described in Table 3.

A stream classification system, developed by the Center for Watershed Protection in Elliott City, Maryland, evaluates a watershed according to the degree of impervious surfaces. The classification system is comprised of three components. These are:

(1)Sensitive watersheds with less than or equal to 10% imperviousness. They are represented by a stable channel, excellent biodiversity, and excellent water quality.

(2)Impacted watersheds with impervious surfaces between 10-25%. They arerepresented by a channel becoming unstable, fair to good biodiversity, and fair to good water quality.

(3)Non-supporting watersheds with impervious surface areas greater than 25% imperviousness. They are represented by poor to no biodiversity and poor water quality.

Impervious surfaces significantly influence the quantity and quality of stormwater runoff from urban and suburban areas. Runoff from lands modified by human activities can harm surface water resources in two ways: (1) by changing the natural hydrologic patterns; and (2) by elevating pollutant concentrations and loadings. Stormwater runoff may contain or mobilize high levels of contaminants, such as sediment, suspended solids, nutrients, heavy metals, pathogens, toxins, oxygen-demanding substances, and floatables. (From USEPA web site - http://www.epa.gov/owmitnet/pipes/3.htm).

The U.S. Environmental Protection Agency is proposing amendments to Section 401 of the Clean Water Act for municipalities with populations exceeding 10,000 persons and population densities greater than 1,000 persons per square mile. In Ocean County, there are 25 municipalities that fall into this category. (Refer to Appendix B for a breakdown of the 33 municipalities in Ocean County by population, land area, and population density.) Municipalities within these categories will be required to obtain National Pollutant Discharge Elimination System (NPDES) permits.

The proposed regulations will require the implementation and development of a cost-effective operation and maintenance program with the ultimate goal of preventing and reducing pollutant runoff from municipal operations. The minimum operation and maintenance components for inclusion in the program are described as follows:

(1)Maintenance activities, maintenance schedules, and long-term

inspection procedures for structural and other stormwater controls

to reduce floatables and other pollutants discharged from the

storm sewers.

(2)Controls for reducing or eliminating the discharge of pollutants from

streets, roads, highways, municipal parking lots, maintenance and

storage yards, and waste transfer stations.

(3)Procedures for properly disposing of waste removed from streets, roads, highways, municipal parking lots, maintenance and storage yards, and waste transfer stations.

(4.)Ways to insure that new flood management projects assess the

impacts on water quality and examine existing projects for incorporating

additional water quality protection devices or practices.

Rain drops fall on the soil in a generally vertical direction, and flowing water moves horizontally over the landscape creating runoff. The roughness and slope of the landscape greatly influence the amount and speed of the runoff. When the rate of the runoff exceeds the absorptive capacity of the soil and land cover, runoff occurs.

Ocean County has seen the installation of more than 1000 stormwater basins to help reduce the peak discharge from changing land uses. These basins should be viewed for multiple purposes, such as, water quality and ground water recharge. There is a need to encourage stormwater basin design, which minimizes natural processes.

Ocean County soils are very sandy. These well-drained soils, in a natural wooded condition, normally do not exhibit any runoff. Much of the rainfall is absorbed in the leaf duff and soaks into the sandy soil. It then reaches shallow ground water, which supplies base flow for the streams in the county.

During colonial times, the streams feeding Barnegat Bay were shallow and narrow. Vegetation impinged on the stream banks, and sponge-like soils soaked up nearly all of the rainfall. Only during a hurricane or other intense storm event did runoff occur.

Sandy soils and natural riparian buffers act to slow runoff and maintain flood plains. These natural buffer zones lessen the amount of pollution entering streams in the watershed. Tree and shrub roots hold the stream banks in place, thus minimizing erosion and sedimentation. The vegetation also slows the flow of the runoff, enabling sediment to settle and runoff to percolate and filter through the soil. This process recharges the ground water.

Currently, vegetation has been removed along many of the stream banks, and subwatersheds have been altered by changes in the landscape. The addition of impervious surfaces, together with compacted soils, has greatly increased the amount of runoff. As pitch pine and other native vegetation are stripped away from the land, pervious areas, which naturally act like sponges to recharge ground water and filter pollutants, are converted to impervious surfaces. Pollutants come into contact with the impervious surfaces and are quickly washed off of the land into the streams. With the construction of roads, roofs, and parking lots, rainwater that once recharged the aquifers of Ocean County, instead becomes runoff. These changes in the land use increase the demand for water and, reduce with every home, the ability of the watershed to recharge its ground water. As a result, many private wells must be drilled deeper to compensate for the reduced ground water recharge.

Nutrients are necessary for vegetative growth. The primary nutrients that can impact the environment are nitrogen, phosphorus, silicon, and potassium. The nutrients in residential and commercial areas are usually applied in inorganic or organic fertilizers. When applied in excess of the needs of lawns or gardens, nutrients may run off into surface waters or leach into ground water.

In the case of deep on-lot sewage disposal systems, nutrients from human wastes may enter the shallow ground water table. Nitrates in excess of 10 mg/l can cause human infant methemolglobinemia (blue baby syndrome), which can be fatal. Adults are less susceptible to this disorder.

Nutrient availability in excess of plant growth needs can lead to eutrophication problems. Nutrient enrichment often results in excessive aquatic plant growth that interferes with recreational activities, and in extreme cases can seriously deplete dissolved oxygen concentrations. Hypoxia and anoxia greatly impact the quality and diversity of habitats for benthic organisms and finfish populations.

Fertilizers usually contain nitrogen, phosphate, and potash. Nitrogen is the most important lawn nutrient, but it can contaminate groundwater with nitrate. Phosphate can contaminate streams, rivers, ponds, and lakes stimulating excessive growth of algae and rooted aquatic plants. Chloride is often combined with potassium in potash, which can also contaminate ground water.

Lawn fertilizers contribute 5-10 kg/ha/yr of nitrates from typical residential lots (Palone and Todd, 1997). In the Barnegat Bay watershed, there is the potential for 286,580 to 573,200 kg/ha/yr of nitrate washoff from lawns (OCPD, 1998). This washoff rate was determined by taking the 1997 residential population and dividing it by 3.5 persons per household and then multiplying by the unit figures for nitrates from lawn fertilizers.

Residents should compost their yard wastes to minimize the release of nutrients to waterbodies. In addition, residents should test their soil so that they will apply only the fertilizer needed. It has been reported that less than 20% of the residents in Ocean County test their soils (Granden, 1993)

The estimated septic effluent flow amounts to 1,328,830 gal/d (5,029,622 l/d) of septic effluent that recharges the ground water table within Ocean County. This value is calculated by multiplying the total of 8437 septic systems within the Barnegat Bay watershed (Appendix C) by 3.5 persons per system, and 45 gal (170 l) per person per day (not including garbage grinders) (USEPA, 1980). Nitrate loadings from septic systems are estimated as follows: (1) 1/2 ac (0.2 ha) lot septic systems (= 25-35 kg/ha/yr); (2) ≥ 2 acre (0.81 ha) lot septic systems (= 14 kg/ha/yr) (Palone and Todd, 1997). Hence, with the 8437 septic systems, the range of nitrates for 1/2 ac (0.2 ha) lots can range from 216,930 to 295,300 kg/ha/yr.

Sediments from the watershed (sand, silt, clay and organic matter) are transported into the Barnegat Bay-Little Egg Harbor estuary and distributed from one point to another by waves and currents generated by winds and storms. Some of the sediment sources include stream banks, construction sites, and agricultural lands. Bottom sediments in the estuary provide habitat for benthic flora and fauna. Fine-grained sediments (clay and silt) tend to retain contaminants that enter the estuary (e.g., trace metals, hydrocarbons, and other toxic chemicals). (Refer to Appendix E for water quality and suspended solids information from Ocean County demonstration projects.)

Human wastes are either treated in on-lot sewage disposal systems or community sewage treatment plants. In on-lot sewage disposal systems, the wastes are usually subjected to primary treatment and effluent disposal. This means that the solids settle out of the sewage in a septic tank, and the liquid effluent is filtered into the ground via a leach field. In some cases, an aerobic treatment tank is used, which removes the sediment and some nutrients. The effluent is then discharged to a leach field.

Community sewage disposal systems involve a series of collection pipes to a sewage treatment plant. At the plant, the sewage is subjected to primary treatment (sediment removal), secondary treatment (reduction of the biochemical oxygen demand, and/or tertiary treatment (reduction of nutrients and other toxics). The treated effluent is then discharged to a stream, the estuary, or is applied to the land.

Stormwater runoff can also transport petroleum products into streams and the estuary via storm drains. These products are residues from vehicles and other motorized equipment. Petroleum hydrocarbons derived from boats, marinas, and land contributions can be assimilated by phytoplankton without necessarily inhibiting their growth. Once contaminated phytoplankton are ingested by herbivorous zooplankton and shellfish, food-chain accumulation may occur and, where there is chronic exposure, the ability of affected organisms to reproduce and feed may be inhibited. (Refer to Appendix E for hydrocarbons discharged from demonstration sites in Ocean County.)

The malfunctioning of on-lot sewage disposal systems often releases wastes. When this occurs, pathogens may enter the surface waters (streams, ponds, lakes, and the estuary) and the shallow groundwater table of an area. Appendix C contains a list of the septic systems in Ocean County.

The concentrations of coliform bacteria are used to set standards that regulate surface water quality, swimming, and shellfish harvesting locations within the Barnegat Bay-Little Egg Harbor estuary. Total and fecal coliform bacteria originate from a multitude of sources, including failing on-lot sewage disposal systems, domestic animals (pets), wildlife, and waste disposal from boats and marinas. Appendix D provides a list of the number of licensed dogs and cats in Ocean County.

Beaches have been historically sampled in the Barnegat Bay-Little Egg Harbor estuary for total and fecal coliform bacteria to assess water quality conditions. Total coliform data have been collected, particularly in shellfish harvesting areas, from 1976 though the present. A major concern is for shellfish harvesting areas in the upper portions of the estuary where tidal influences and dilution from the ocean are minimal.

Shellfish growing area classifications have been established based on total and fecal coliform bacteria levels and nonpoint pollution sources. These classifications are:

1. Approved Areas: Waters that have been approved for shellfish harvesting.

2. Prohibited Areas: Areas where shellfish harvesting is not permitted under

any circumstance.

3. Special Restricted Areas: Areas from which shellfish must be purified before

human consumption. Applications for removal will be considered for

transplant, relaying, and controlled purification.

4. Seasonal Area: Waters that are condemned and opened for the harvest of

shellfish each year on either January 1 or November 1. All Seasonal Areas

close April 30 of each year.

5. Condemned Areas: Waters that are prohibited, special restricted, and seasonal areas.

During the sewage treatment process, sludges are produced from sedimentation. This sludge usually contains heavy metals. These contaminants may pose a problem if the wastes are considered for land-use disposal on edible crops. Appendix E provides information on fecal coliforms at Ocean County demonstration sites.

There are household hazardous products that can present a problem if stored in sufficient quantities, and a spill occurs near a storm drain, stream, well, or pond. There are numerous household materials, namely, trash, clothing, and fabric care products, hobby and recreation products, building/wood cleaner and repair products, pesticides, and vehicle maintenance chemicals that may accumulate in streams, lakes, or ponds via storm drains or direct discharge. Lead contamination usually derives from lead-based paint, water from lead pipes, or lead solder. The soil may be contaminated from prior manufacturing activities.

Household and commercial waste control reqirees wise management of trash and other solid waste. Many municipalities have adopted curb-side trash separation programs for glass, papers, cans, cardboard, and other reusables or recyclables.

Over the years, numerous septic tank cleaners have been used. Such chemicals include TCE (Trichloroethylene) and other chlorinated hydrocarbons. Some of these chemicals have been found in surficial ground water.

Pesticides include herbicides, insecticides, fungicides, rodenticides, nematocides, algicides, and bactericides used to control pests to human habitation. The most common are the herbicides which prevent weed growth in lawns. Other pesticides are used to control termites, ants, roaches, rodents, and bacteria in swimming pools. Wise use is recommended. The State of New Jersey provides certifications to applicators to ensure that the proper dose is applied to preclude overexposure to humans. Many counties have instituted hazardous waste collection days for receipt of containers of chemicals (e.g., paint, pesticides, cleaners, and other products) at designated locations.

Many of the soils in Ocean County have been disturbed through cut-and-fill operations (Smith, 1999). Site specific analyses are needed for interpreting residential and commercial applications, such as on-lot sewage disposal systems, basement drainage, surface runoff considerations, and fertilizer and pesticide applications. A runoff curve number analysis is currently being evaluated for the soils in the county both prior to land development and after land development.

With the support of the United States Department of Agriculture and the Natural Resources Conservation Service, the Ocean County Soil Conservation District began conducting soil bulk density tests. Soil bulk density is the ratio of dried soil (mass) to its bulk volume, including the volume of particles and the pore space between the particles. Preliminary findings have shown that typically undisturbed sandy soil has a bulk density of 1.08 g/cc. Following grading activities, bulk densities have ranged from 1.59 to 1.97 g/cc. Regardless of texture, any soil layer with a bulk density greater than 1.65 g/cc is considered root restrictive. Hence, there is an increased amount of surface runoff and surface ponding after storm events. (Friedman, 1999; Smith, 1999)

Ocean County has the greatest number of private wells of any county in New Jersey. Most of these wells are in the Kirkwood-Cohansey aquifer. This aquifer is highly vulnerable to activities on the landscape.

When wells are used in conjunction with residential or commercial irrigation systems, backflow prevention devices are effective in preventing cross-connections with the potable water system. This is especially a concern if fertilizers and/or herbicides are incorporated into an irrigation system. Abandoned private wells are a source of pollution if not properly sealed.

Potential sources of groundwater contamination, which may be present near a home site, include septic tanks, animal waste, pesticides, fertilizers, fuel storage tanks, household chemicals, and used motor oil.

There are six basic principles of well head protection. These are:

1.Proper well siting

2.Proper well construction

3.Keeping contaminants away from the well

4.Use of backflow prevention

5.Sealing abandoned wells

Testing well water

 

Agricultural Nonpoint Source Pollution: Ground Water and Surface Water

To characterize agriculture in the Barnegat Bay watershed, it is necessary to look at past activities and trends, as well as the present and future condition, as they may impact water quality. The information in this section is based on agricultural data compiled for Ocean County and, since nearly all of Ocean County is within the watershed, it provides the most comprehensive way of characterizing the watershed. Prior to 1970, agriculture within the Barnegat Bay watershed was characterized primarily by chicken farms and small cranberry bogs. Table 4 shows the trends in the number of farms and horses in Ocean County since 1930.

Historic impacts of agriculture on current water quality may be significant due to the long residence time of groundwater before it appears as surface water contributions to surface water tributaries and the estuary. Figure 1 shows the extent of agricultural activity, particularly poultry farming, in Ocean County in 1954. Figure 2 illustrates the groundwater residence times for areas of the watershed. These two maps together provide useful information on the nature and degree of historic agricultural impacts on present surface water quality. Potential pollutants associated with agriculture in the Barnegat Bay watershed include nutrients, sediment, pathogens and pesticides.

The 1987 State Equine Survey showed that Ocean County had 1,500 horses, with 1,300 ac (520 ha) dedicated to horse farming (New Jersey, 1987). According to the 1996 State Equine Survey, Ocean County had 1,300 horses on 3,500 ac (1,400 ha). The number of horses appears to have declined considerably when this information is compared to the 1997 agriculture census (Table 5). However, it should be noted that the 1996 State Equine Survey indicates that the total number of horses in the county, including those residing on farmettes, is approximatley twice the number shown for 1997.

The poultry industry has been a significant part of agriculture in Ocean County for many years. Chicken farms were numerous and produced substantial quantities of both eggs and poultry for the New York and Philadelphia markets during the 20^th Century. These operations also produced large quantities of manure high in nitrogen and phosphorus, which was repeatedly spread on sandy, porous soils within the watershed.

Data in Table 6 indicate the trends in the number of farms and number of chickens in Ocean County since 1930. Figure 3 shows the number of chickens that were four months old and older by municipality in Ocean County during 1954. Dover Township had the greatest number of poultry (1,173,908) followed by Jackson (958,411) and Brick/Lakewood (372,487). As is evident in Table 6, the peak year for poultry production in the county was 1969.

Ground water that originates near a stream follows short flow paths that require short travel times prior to seeping into the stream, whereas ground water that originates from distant sources follows relatively long flow paths that require longer travel times prior to entering the stream flow (Modica, 1999). Concentrations of total nitrogen and nitrate in Wrangel Brook, a tributary of the Toms River, were found to be greater during base flow than during storm flow. These nutrients are thought to derive largely from nitrate used in fertilizers for high-maintenance lawns and from agricultural runoff of poultry farms located in the basin almost fifty years ago (1950s) (Hunchak-Kariouk, 1999). As a result, a significant amount of nitrate impacts to the estuary probably reflect poultry farming conditions decades ago. In relative terms, the decline in the number of poultry farms in Ocean County in recent years suggests that they probably have little ongoing impact on water quality, especially when compared to potential contributions from other sources such as lawn fertilization, septic systems, and wildlife.

Other livestock enterprises also exist in Ocean County, but their impacts on water resources are likely to be less significant than those resulting from horse and poultry farming. Table 7 shows the number of farms and animals for cattle and calves, hogs and pigs, and sheep and lambs in Ocean County. Table 8 indicates the number of farms and number of turkeys and other poultry in the county.

The cranberry industry has played an important role in Ocean County agriculture over the years. Acreage devoted to bogs was minimal and the production systems were not high input. Consequently, it is unlikely that water quality was adversely impacted by this type of farming. Table 9 reveals the number of farms and acres in cranberry production in Ocean County since 1930.

Nutrients are necessary for crop production. Nitrogen and phosphorus are the primary nutrients that can impact the environment. They may originate from inorganic fertilizers, from land-applied animal wastes, or from other organic residuals such as sludge. When applied in excess of crop needs, nutrients run off into surface waters resulting in excessive aquatic plant growth and toxicity to certain fish species. Accelerated plant growth in aquatic systems often interferes with recreational activities and can impact some habitats. Severely depleted dissolved oxygen levels (i.e, hypoxia or anoxia) associated with excessive plant growth reduce the quality of habitats for fish, invertebrates, and other organisms.

Nitrogen losses from manure may be high. For example, volatilization can account for up to a 70% loss of nitrogen from manure as ammonia gas. Nitrogen lost in this way is not available for plant use, runoff, or entry into groundwater. As a result, farmers interested in reducing fertilizer costs and lessening the potential for ground water and surface water impacts inject or incorporate manure into the soil. This keeps volatization losses to less than 5%. If manure is applied in the fall or winter when there is no crop to actively assimilate the nitrogen, particularly on sandy soils, as much as 50% of the nitrogen can be lost through denitrification and leaching into ground water. Denitrification losses can be as high as 20% of manure nitrogen.

Nitrate-nitrogen is highly mobile and thus can be leached through the crop root zone to ground water, especially on sandy soils. Nitrate concentrations > 10 mg/l can cause human infant methemoglobinemia, which can be fatal. It can also result in cattle abortion and other livestock disorders. Nitrogen is not routinely evaluated in typical soil tests; however, surveys in this watershed and others in New Jersey show that no more than 10 to 25% of the public test their soil before applying lawn and garden fertilizer (Grandin, 1993; OCSCD, 1995-1997). These same surveys reveal that few lawn service companies test soils prior to fertilizing lawns.

Nitrate is a major constituent in ground water sampled in some residential areas (Watt et al., 1994). Ground water sampled in a number of wells located near streams has exhibited elevated nitrate concentrations. Three samples, collected on Maple Root Branch, Union Branch, and Old Hurricane Brook, exceeded the 10 mg/L MCL (Maximum Contaminant Loading). Median concentrations of total nitrogen were generally highest at sites on the Toms River, downstream on Davenport Branch, Wrangel Brook, and Long Swamp Creek. Figure 4 shows the extent of cropland harvested acres recorded in 1954. Table 10 lists the agricultural land use over time. In 1997, the Census of Agriculture revealed that fertilizer use was reported for 106 farms on 2925 ac (1170) of land in Ocean County (USDA, 1993-1997). Moser (1997) found that ground water contributed about three times as much nitrogen to the northern portion of the Barnegat Bay-Little Egg Harbor estuary as to the southern portion, although the overall contribution is relatively small compared to atmospheric and river nitrogen inputs.

Phosphorus occurs in particulate and dissolved forms in runoff from the watershed. Particulate phosphorus is sorbed to mineral and organic sediment as it moves with the runoff. In general, particulate phosphorus constitutes most (75-90%) of the phosphorus transported in runoff from cultivated land. Dissolved phosphorus comprises a larger portion of the total phosphorus in runoff from non-cultivated lands, such as pastures, fields with reduced tillage, and lawns. Although dissolved phosphorus is 100% bioavailable for plant growth, particulate phosphorus is only 10 to 90% bioavailable because of its association with mineral and organic compounds (USDA, 1994).

Sharpley (1997) noted that all soils do not contribute equally to phosphorus export from watersheds or have the same potential to transport phosphorus to runoff. In their studies, Coale and Olear (1996) observed that soil test phosphorus levels did not accurately predict total dissolved phosphorus. However, they noted that, in all cases studied, soil test phosphorus levels were significantly related to an increase in total dissolved phosphorus. Soil testing is currently the best management tool available to ensure that soils do not become overloaded with phosphorus, which increases the likelihood of their contribution to pollution in surface waters downstream (Norfleet et al., 1996).

Soil test results for the Barnegat Bay region are good indicators of phosphorus availability for movement into surface and ground water of the watershed. A review of soil test results in the Great Swamp watershed in Morris and Somerset Counties found that a large proportion of soil tests had high or very high phosphorus levels (Westfall, 1993). Figure 5 shows the summary of statewide soil test results for phosphorus in soils. A review of the 1997 soil test results from the Rutgers Soil Test Laboratory indicates that 83.9% of soil samples received (801 samples) from around the state had high or very high phosphorus levels.

There are several options available to more effectively manage phosphorus. These include base fertilizer application and placement as well as agronomic considerations. Where soil phosphorus tests are high, applications may be eliminated. Among the agricultural practices which can minimize runoff of phosphorus are subsurface application, conservation tillage, buffer and filter strips, crop rotations with legumes, terracing, contouring, and the use of cover crops (USDA, 1997a). Because agriculture is not a significant land use within the watershed, total phosphorus levels are more likely based on particulate phosphorus derived from barren areas, road ditches, and construction sites, as well as dissolved phosphorus originating from poorly managed lawns.

Sediments can significantly affect environmental conditions in aquatic systems. For example, sediments in suspension reduce the amount of sunlight available to benthic plants and may clog the gills of benthic animals. Nutrients and chemical contaminants sorbed to sediments often adversely impact water quality. In addition, turbidity commonly creates conditions that are undesirable for swimming, fishing, and other recreational pursuits.

The 1979 State Erosion, Sediment and Animal Waste Study (SESAW) found that on 7,300 ac (2,920 ha) of cropland in the Manasquan-Metedeconk River watershed the average annual sheet and rill erosion rate amounted to 4.8 tons per acre (USDA, 1986). There was an average annual sheet and rill erosion rate of 5.2 tons per acre on 6,797 ac (2,719 ha) of cultivated cropland. No other information was gathered for the remainder of the Barnegat Bay watershed. The National Resource Inventory, as currently configured, does not provide soil erosion information that is statistically significant at the watershed level. In New Jersey, there are exploratory efforts now underway to determine the level of interest in increasing the sample size of the National Resource Inventory to make it more useful for measurements of the "state of the land" at the watershed level.

Pathogens in the form of bacteria and viruses can seriously impact the health of livestock, wildlife, aquatic organisms, and humans. Organic wastes primarily derived from livestock (e.g., dairy cattle, chickens, hogs and horses) are important sources of pathogens (as well as nutrients) to receiving waters in some regions. Wastes from sludges and food processing residuals are also potentially significant sources.

Pesticides primarily include herbicides, insecticides, and fungicides. Pesticides, while often providing a substantial benefit to crop production, may impact the environment by adversely affecting non-target organisms. Negative effects are acute or chronic toxicity and accumulation in organism tissues. One study (Wauchope, 1978) has indicated that generally less than 0.5% of the total pesticides applied in the field reach receiving waters. Most pesticides leave the field in soluble form in runoff and not attached to sediment.

Arsenical pesticides were historically applied to cropland, turf, and golf courses between 1900 and 1980. Approximately 100,000 lb (450,000 kg) of arsenic were applied cumulatively in Ocean County (Murphy and Aucott, 1998). Arsenic is not considered to be highly mobile in soils, but its downward migration has been shown to be greater in a sandy soil, such as those prevalent in the Barnegat Bay-Little Egg Harbor watershed, than in a clay loam (Elving et al., 1994).

The 1997 Census of Agriculture shows general agricultural chemical use for control of insects, disease, and weeds (Table 11). Pesticide use in the Barnegat Bay watershed (Managment Area 13) is relatively low when compared to the other watershed management areas in the state (Brown, 1999). Figure 6 illustrates pesticide use by watershed management area.

Recent land-use data show that agriculture comprises ~3% of the Barnegat Bay watershed area (USDA, 1997). Farmland assessment data for 1993 (Adams, 1993) and 1996 (Adams, 1997) reveal that there was 5,602 ac (2,241 ha) and 5,488 ac (2,195 ha) of harvested cropland area, respectively, in Ocean County. These harvested areas were located in Barnegat, Berkeley, Brick, Dover, Eagleswood, Jackson, Lacey, Lakewood, Manchester, Ocean, Plumsted and Stafford Townships. Approximately one-half of the acreage is devoted to crop production (grain, vegetables, nursery) and the remainder to horse farms. More than 75% of the agricultural use occurs within the Toms River and Metedeconk River subwatersheds. Table 12 lists the land-use trends in Ocean County over time.

One indicator of the extent of future agricultural land use in Ocean County is farmland preservation. Approximately 1,750 ac (700 ha) of land have been preserved in Plumstead, which is located in the Delaware River Basin. No acres have been permanently preserved in the Barnegat Bay watershed (McKeon, 1999).

Production agricultural fields within the watershed are generally flat and well drained. There is a minimal hazard of runoff and soil erosion under well-managed conditions. The row crops produce only small amounts of sediment, nutrients, and pesticides, with delivery to surface waters limited to headwater drainages. Because of the soils, topography, and limited agricultural acreage within the watershed, significant impacts to the estuary from production agriculture are unlikely.

Livestock acreage, predominantly consisting of small equestrian enterprises, has a slightly greater potential for impact. "Hobby" and commercial horse boarding and riding operations generally share two common characteristics: (1) exercise lots and paddocks that are barren (no vegetation) and possess highly compacted soils; and (2) improper stockpiling of manure. The soils, being bare and relatively impervious, are generators of sediment and excess runoff. The runoff picks up animal wastes in the lot and also may flow through manure stockpiles at the periphery of the property, releasing phosphorus, nitrogen, organic matter, and pathogens to surface waters.

Resource management systems should be developed for all significant horse operations in the Barnegat Bay watershed, and a manure management cooperative should be developed with surrounding row crop farmers and mushroom growers in order to utilize the waste as a resource. These resource management systems would entail, to the greatest extent practical, the exclusion of clean water from roofs and upper lying land from animal use areas. The nature and extent of current riparian buffers adjacent to agricultural and other land uses within the watershed should be inventoried. Riparian buffers should be planted or maintained where agricultural cropland and animal-use areas interface with streams and water bodies.

This section describes the wastewater collection and treatment facilities within the Barnegat Bay watershed, including a historical look at the situation prior to the implementation of the Ocean County Sewerage (Utilities) Authority regional wastewater treatment system. Since about 1980, all discharges of treated wastewater to the Barnegat Bay watershed have ceased. There are no point source discharges of wastewater into the watershed. For the purposes of this section, stormwater outlets are considered nonpoint sources.

In 1970, the Ocean County Utilities Authority (OCUA) was created to regionalize the treatment of wastewaters generated by residential, commercial, and industrial sources within the service area. Prior to the creation of the regional authority, there were 46 small to medium sized treatment plants operated by municipalities, developers, private sewer companies, and one industry within the OCUA service area. The current OCUA service area also includes certain areas outside the boundaries of the Barnegat Bay watershed, primarily those within the Manasquan River watershed. Prior to the implementation of the regional wastewater system, extensive areas of the watershed were single-family homes utilizing septic systems, many of which were adjacent to Barnegat Bay. Figure 7 shows the locations of the wastewater treatment plants in 1972. Except for wastewater treatment plants on the barrier beaches which discharged into the Atlantic Ocean, all discharges prior to 1972 were to the Barnegat Bay watershed either directly or via streams and rivers. The treatment plants discharging to the Barnegat Bay watershed were generally secondary or higher level treatment with the exception of the U.S. Navy plant at Lakehurst (0.275 MGD; 1.040 MLD) and the Forked River State Marina (0.006 MGD; 0.085 MLD) which provided only primary sewage treatment.

Table 13 indicates the average 1971 summer and winter flows, as well as the level of treatment, and the receiving waters for the treatment plants in the watershed. Based on this data, during the summer of 1971, an average of 6.866 million gallons (25.988 million liters) of effluent per day was discharged to the Barnegat Bay watershed. An additional 13.608 million gallons (52.731 million liters) of effluent per day were discharged to the Atlantic Ocean during the 1971 summer season, which included 5.0 million gallons (18.9 million liters) per day of effluent from the outfall of the Ciba-Geigy chemical plant in Toms River.

Figure 7 indicates the location of the three existing OCUA Regional Wastewater Treatment Facilities and their outfall lines. These outfall lines include 1 mi (1.6 km) of solid pipe that connect to diffuser sections averaging 1400 ft (426.7 m) in length that mix the effluent with the ambient ocean water.

In 1985, the OCUA service area was expanded to include the Manasquan River Regional Sewerage Authority system which encompasses areas of southern Monmouth County that are not part of the Barnegat Bay watershed. During 1998, OCUA treated an average of approximately 52 million gallons (196.8 million liters) per day. However, during rainfall events this volume was significantly higher because some municipal collection systems allow infiltration and inflow of rainwater. The average daily flow increases to approximately 60 million gallons (227.1 million liters) per day during the summer months.

Unlike most estuaries, the Barnegat Bay-Little Egg Harbor stuary does not have any direct input of pollutants from wastewater treatment plants. Nutrient loading and fecal coliform contamination is associated with nonpoint sources of pollution. One concern that has been raised recently is the consumptive use of freshwater by wastewater treatment plants located in the watershed. Because the treatment plant effluents discharge directly into the ocean, significant volumes of freshwater are being exported out of the watershed without any beneficial re-use of the water. The possible impacts to the watershed of this substantial freshwater export is not fully understood. However, the potential for reclamation of wastewater treatment plant effluent to make water available in the watershed for irrigation, cooling water, or groundwater recharge should be investigated.

Septic systems, underground storage tanks, and landfills are all potential sources of ground water pollution in the watershed. Cohansey sands, which comprise most of the Barnegat Bay watershed, do little to absorb or breakdown contaminants, which may be intentionally deposited in an individual subsurface sewage disposal systems (i.e., septic system). The relatively short distance between the bottom of the septic system and the groundwater has led to episodes of ground water contamination.

Homeowners, ignorant of the consequences, have irresponsibly flushed paint thinners, turpentine, household degreasers, cleaners, and detergents into septic systems.  Because of the inability of Cohansey sands to filter these chemicals, the contaminants percolate through the soils until they reach the water table where they can ultimately empty into the estuary.

Underground storage tanks, whether housing gasoline or home heating oil, have often leaked their contents into the Cohansey sands. While tanks with a capacity of 2,000 gallons (7,570 liters) are required to be periodically tested and to have leak detection measures, there are no such requirements for the homeowner. These tanks also have been the focus of several investigations of groundwater contamination.

Underground storage tanks all have the potential leak contaminants into the ground water over time. Within Ocean County, a database of underground storage tanks has been developed by the New Jersey Department of Environmental Protection. A number of entities have been identified that contain underground storage tanks (Table 14). Each of these entities may have several tanks. However it is difficult to keep this database up to date, because the tanks are replaced or removed periodically. This list merely indicates that there are a significant number of underground storage tanks within Ocean County, New Jersey. Specific information on the various tanks, such as the material construction, they types of materials stored, and the service life will require an inquiry with the entity involved and/or the New Jersey Department of Environmental Protection.

Sanitary landfills that were constructed in the past were little more than dump sites. They were constructed without liners or caps so that rainwater percolated through the garbage, concentrated sewage and chemical wastes and, ultimately drained down into underlying soils and ground water. Landfills in Ocean County remain potential sources of ground water contamination.

The fate of contaminants in ground water and whether or not contaminants reach the estuary depends on many factors, notably aquifer characteristics and the chemical characteristics of the contaminant. Aquifer characteristics include whether the aquifer is confined or unconfined, the composition of aquifer material, and the composition of the soils. The amount and quality of water recharging the aquifer is also a factor. Ground water is most vulnerable to contamination in areas that have well-drained (loamy, sandy) soils that are underlain by unconsolidated sand and gravel aquifers where no protective confining layer exists. Fine-grained materials, such as clays and silts, however, act as confining layers that restrict the downward movement of water and contaminants. Clays and silts also enhance chemical adsorption of some contaminants, thereby decreasing the availability of contaminants which could moved through the soils to ground water.

Chemical characteristics are specific to each compound or element, and chemical properties may vary due to chemical composition and form. Water solubility, soil and aquifer sorption, and in the case of pesticides and VOCs, persistence, diffusion, and dispersion, are some of the chemical characteristics that may determine how long and how far a contaminant will travel in the ground water system. For example, a water-soluble organic compound that does not degrade either by chemical, physical, or microbial mechanisms and does not easily sorb to soil or aquifer material, can be very mobile and may be carried a significant distance in a shallow ground water system. Inorganic and organic compounds may be exposed to varying chemical conditions that may change along flowpaths in the aquifer; changes can include the transition from aerobic to anaerobic conditions, changes in pH, and temperature, which can alter the form of the compound and thus change its chemical properties.

For the time period water years 1987-97, surface-water quality data for nutrients (total phosphorus, nitrogen, nitrate plus nitrite, ammonia plus organic nitrogen, and ammonia) and sulfate were available for 24 stations located on nine of the 12 rivers and streams which flow directly into the Barnegat Bay-Little Egg Harbor (BBLEH) estuary or the Toms River embayment. Surface water quality data were available from the Pinelands Commission, USGS, and the Leeds Point laboratory of the NJDEP for stations on the Toms River, Wrangel Brook, Jakes Branch, Long Swamp Creek, Cedar Creek, Forked River, and Oyster, Mill, and Westecunk Creeks. No water quality data were available for the Cedar Run, or Tuckerton Creek. Some data were available for the Metedeconk River; but for this analysis, these data were insufficient and not comparable to data from the other stations. The station on McDonalds Branch is not within the BBLEH watershed, but was included in the analysis because it represents undeveloped areas within the New Jersey Coastal Plain. Loads for 10 stations on 9 streams draining 266 mi² (691.6 km²) of the 419 mi² (1089.4 km²) BBLEH watershed were calculated for a 6-month, high-flow, nongrowing-season period and a 6-month, low-flow, growing-season period by using the median of the concentrations measured on dates of high and low flows, respectively, and stream flows for the 25% and 75% flow durations, respectively.

Two landscape categories were defined for the watershed based on the percent of urban land cover in 25 watersheds of the New Jersey Coastal Plain. Stations with less than ~10% of urban land cover were designated as Landscape Category I, and stations with more than ~10% of urban land cover were designated as Landscape Category II. Yields (lb/mi²/yr) and loads (lb/yr) for each landscape category were determined as the median of the yields and loads for the stations in each category. In the watershed, 153 mi² (397.8 km²) of the 419 mi² (1089.4 km²) contributing area to the BBLEH were unmonitored. The percent land use was determined for the unmonitored areas downstream of each sampling station in each river basin. These data were used to classify each unmonitored area as either Landscape I or II. The appropriate yields for each landscape were multiplied by the area to determine the loads for each unmonitored area. Annual loads (lb/yr) of total phosphorus, nitrogen, nitrate plus nitrite, ammonia plus organic nitrogen, and ammonia and sulfate to the BBLEH from surface runoff were determined as the sum of the loads from all monitored and unmonitored areas of the watershed.

The load contributed by each river to the BBLEH is dependent on the size of and type of land cover in the basin. Larger basins will contribute larger loads because they have larger annual stream flows; however, the yield from each river basin will depend on the intensity and type of land cover in the basin. The largest basins of the BBLEH watershed are the Toms River (31%, 128 mi²; 338 km²), Metedeconk River (18%, 74.5 mi²; 193.7 km²), Cedar Creek (13%, 56 mi²; 145.6 km²), and Wrangel Brook (8%, 34.6 mi²; 90 km²) basins. Urban land cover accounts for 19% of the entire BBLEH watershed. The Long Swamp Creek basin has the greatest percentage of urban land cover (55%). Only about two-thirds of the contributing area (15% urban land cover) to the BBLEH was monitored. Most development in the BBLEH watershed is in the coastal, unmonitored areas (28% urban land cover). The magnitude of loads and yields from unmonitored areas compared to the loads and yields from the rest of the basin varies by constituent.

Median total phosphorus (TP) concentrations at the 25 stations during all flows ranged from less than 0.01 to 0.06 mg/l (Table 15). At the 12 stations categorized as Landscape I, concentrations during high and low flows were not significantly different; the median TP concentrations during all flows were 0.01 mg/l or less. For the 13 stations categorized as Landscape II, the median TP concentrations during all flows ranged from 0.01 mg/l to 0.08 mg/l; 7 stations had median concentrations of 0.02 mg/l or greater. Seven of the Landscape II stations had higher TP concentrations during low flows than during high flows. Nonpoint storm runoff does not seem to be a significant source of TP to the bay because concentrations were low, especially in areas of slight urban land cover, and either the same or greater during low flow than high flow in areas of moderate to high urban land cover.

The annual load of TP to the bay from surface water discharge was estimated to be 23,000 lb/yr (10,350 kg/yr) (Figure 8 and Table 17). The highest TP basin loads were from the Toms River (9,000 lb/yr, 4,050 kg/yr, or 41% of the total annual load), Wrangel Brook (2,300 lb/yr, 1,350 kg/yr, or 10%), and Cedar Creek (2,300 lb/yr, 1,350 kg/yr, or 10%) basins. The load from the Metedeconk River was also high (3,800 lb/yr, 1,710 kg/yr, or 17%), but is entirely estimated. The lowest loads were from the Long Swamp Creek (291 lb/yr, 131 kg/yr) and Jakes Branch (313 lb/yr, 141 kg/yr) basins. The load from Cedar Run is also low (289 lb/yr, 131 kg/yr), but is entirely estimated. About one-third of the load derives from the unmonitored coastal areas of the basin. TP load appears to be related to the type of land cover - basins with greater urban land cover have higher concentrations, yields, and loads of TP to the bay.

Median total nitrogen (TN) concentrations at the 25 stations during all flow conditions ranged from 0.18 to 0.99 mg/l (Table 15). At the 12 stations designated as Landscape I, median TN concentrations during all flows ranged from 0.18 to 0.53 mg/l; 8 stations had higher median concentrations during high flow than low flow, and 4 stations had higher median concentrations during low flow than high flow. The overall median concentrations during high, low, and all flows were not significantly different (0.30, 0.29, and 0.27 mg/l, respectively). For the 13 stations designated as Landscape II, median TN concentrations during all flows ranged from 0.35 mg/l to 0.99 mg/l; 3 stations had higher median concentrations during high flow than low flow, and 10 stations had higher median concentrations during low flow than high flow. The overall median concentrations during high, low, and all flows were not significantly different (0.71, 0.79, and 0.80 mg/l, respectively).

TN is a measure of several nitrogen species that can be dissolved or associated with particles. The concentration of TN and the relative importance of nonpoint source constant (ground water) and intermittent (storm runoff) contributions at a site depend on which nitrogen species is predominant. Dissolved species such as nitrite plus nitrate are most likely carried to a stream by ground water and can be in higher concentrations during low flow because the concentration may be diluted during storm flow. Ammonia and organic nitrogen can be associated with particles and can be in larger concentrations during high flow because they may be carried to a stream by surface runoff.

The annual load of TN to the bay from surface water discharge was estimated to be 870,000 lb/yr (Figure 9 and Table 17). The highest TN loads were from the Toms River (340,000 lb/yr or 38%), Wrangel Brook (93,000 lb/yr or 11%), and Cedar Creek (71,000 lb/yr or 8%) basins. The load from the Metedeconk River basin was also high (170,000 lb/yr or 19%), but was entirely estimated. The lowest loads were from the Jakes Branch (10,000 lb/yr) and Long Swamp Creek (4,000 lb/yr) basins; the load from the Cedar Run basin was also low (8,000 lb/yr), but was entirely estimated.

Median total ammonia plus organic nitrogen (TAON) concentrations at the 25 stations during all flow conditions ranged from 0.20 mg/l to 0.74 mg/l (Table 16). At the 12 stations designated as Landscape I, median TAON concentrations during all flows ranged from 0.20 mg/l to 0.52 mg/l; 8 stations had higher median concentrations during high flow than low flow, and 2 stations had higher median concentrations during low flow than high flow. At these stations, the overall median concentration during high flow (0.30 mg/l) was somewhat higher than during low flows (0.23 mg/l); the overall median concentration was 0.25 mg/l. At the 13 stations designated as Landscape II, median TAON concentrations during all flows ranged from 0.23 mg/l to 0.74 mg/l; 9 stations had higher median concentrations during high flow than low flow, and 3 stations had higher median concentrations during low flow than high flow. The overall median concentration during high flows (0.40 mg/l) was higher than during low flows (0.31 mg/l); the overall median concentration during all flows was 0.40 mg/l.

Nonpoint storm runoff is most likely an important source of TAON to the bay because concentrations were higher during high flow than low flow, especially in areas of moderate to high urban land cover. TAON also seems to be related to land use; concentrations tend to be higher in areas with more wetlands.

The median yield of TAON was higher during high flow than low flow at all stations (Table 16). The median TAON yield during high flow was slightly higher for Landscape I stations than for Landscape II stations (790 lb/mi2/yr and 700 lb/mi2/yr, respectively). During low flow, it was slightly higher for Landscape II stations than for Landscape I stations (310 lb/mi2/yr and 270 lb/mi2/yr, respectively). Annual yields for Landscapes I and II were similar (1,110 and 960 lb/mi2/yr, respectively). The basin average TAON yield for the bay watershed was 1,110 lb/mi2/yr (Table 16). The highest TAON yields were from the Mill Creek (1,500 lb/mi2/yr), Oyster Creek (1,300 lb/mi2/yr), and Toms River (1,200 lb/mi2/yr) basins; the lowest yield was from the Long Swamp Creek basin (360 lb/mi2/yr) (Table 16). The yields from the monitored and unmonitored portions of the bay watershed were similar (1,200 and 980 lb/mi2/yr, respectively).

The estimated annual load of TAON to bay from surface water discharge was 460,000 lb/yr (207,000 kg/yr) (Table 17). The highest TAON loads were from the Toms River (160,000 lb/yr, 72,000 kg/yr, or 33%), Cedar Creek (64,000 lb/yr, 28,800 kg/yr, or 14%), and Mill Creek (35,000 lb/yr, 15,750 kg/yr, for 8%) basins. The load from the Metedeconk River basin (72,000 lb/yr, 32,400 kg/yr) was also high. The lowest load was from the Long Swamp Creek (2,400 lb/yr, 1080 kg/yr).

Most ground water within basin discharges to gaged streams that flow into the bay. Some ground water does discharge to small ungaged streams near the coast, and some ground water discharges directly to Barnegat Bay. Loadings for three water quality parameters in ground water that discharges directly to the bay or to small unmonitored tributaries were estimated. Values for three variables were needed to calculate a ground water load, the area contributing to direct ground water discharge to the bay and small tributaries, the ground water recharge rate for that area, and a representative concentration.

The area of ground water basins was determined using GIS. The ground water recharge rate was assumed to be 15 in (38.1 cm) per year. This estimated value is based on the recharge rates reported by Watt and others (1994), adjusted downward to account for a relatively high percentage of urban land use, which generally results in lower rates of recharge.

Parameter concentrations used to calculate loads were the median concentration from ground water samples of the 16 wells of various depths within the contributing recharge areas shown in Figure 12. For calculation purposes, data values reported at less than the method detection limit were assumed to equal that limit. Parameter concentrations and medians are listed in Table 18. These medians compare favorably with medians for the same parameters determined by a National Water Quality Assessment (NAWQA) study of shallow, recently recharged ground water from the Kirkwood-Cohansey aquifer system in urban areas (Stackelberg et al., 1997). The medians they report are 2.6 mg/L for NO2+NO3, 22.9 mg/L for SO4, and 2 mg/L alkalinity as CaCO3. Given that the medians used for the loadings are similar in value and come from ground water samples collected in the same aquifer and within urbanized basins, the medians appear to be comparable and reasonable.

The equation to calculate parameter loadings is:

Loading = (Area) x (Concentration) x (Recharge rate) x (Units conversion factor)

Calculated gross loadings are, respectively: 326,488 lb/yr (146,920 kg/yr) NO2+NO3; 1,959,149 lb/yr (881,617 kg/yr) SO4; and 842,644 lb/yr (379,190 kg/yr) alkalinity as CaCO3. An investigation by the USGS National Water Quality Assessment (NAWQA) Program of nitrate concentration in streams predicted from nitrate concentration in shallow ground water through numerical modeling of ground water flow, found that for three different basins in the New Jersey Coastal Plain, the NO2+NO3 concentration measured in streams was consistently about 40% less than the predicted concentration (Leon Kauffman, U.S. Geological Survey, written communication, 1999). Because this NAWQA study was in the same physiographic province and aquifer system as the Barnegat basin, the NO2+NO3 loading was reduced by 40% to 195,892 lb (88,151 kg), while sulfate and alkalinity were assumed to be conservative.

Yields to the bay from the ground water basins were calculated from the loadings, allowing comparison with yields from the surface water basins and ground water yields from other studies. The yields are 2,024 lb/mi²/yr for NO2+NO3, 20,239 lb/mi²/yr for sulfate and 8,705 lb/mi²/yr for alkalinity as CaCO3.

Contributions of constituents to the bay from rivers can be quantified as concentrations (mass per volume), loads (mass per time) and yields (mass per time per unit area). Some important factors influencing the amount of nutrients that enter the bay via river discharge are: (1) the type of land use in the basin, such as urban/residential, grasslands plus agricultural land, or forested land; (2) the intensity of development in the basin (greater than or less than 10% development); (3) the historical land use in the basin; (4) the relative importance of constant (ground water discharge) and intermittent (runoff) contributions of constituents to the river; (5) the relative importance of ground water discharge to annual stream flow of each river; and 6) the area of contributing drainage (basin size). Concentrations and loads are strongly influenced by stream flow - low concentrations can represent a large instream load during high flow conditions; the magnitude of loads is dependent on the size of the basin – larger loads come from larger basins. Yields, loads normalized to the basin area, are directly comparable between basins because the influences of stream flow and basin size are removed.

The main contributors of water quality constituents to the bay are nonpoint sources in the basin because there are few major permitted point source discharges to the major rivers draining the watershed. Nonpoint source constituents are transported to the bay through ground water, surface water, and direct atmospheric deposition. Nonpoint source contributions of nutrients (total phosphorus, nitrite plus nitrate, and total ammonia plus organic nitrogen) are typically associated with certain types of land cover, such as runoff from agriculture and intensive lawn maintenance. Increased amounts of impervious surfaces and soil compaction during building construction can reduce the infiltration rate of rainfall and increase storm runoff. Forested land and wetlands land cover has greater water retention and less storm runoff than other land covers as a result of ponding and dense vegetation. The presence of some constituents in ground water can be attributed to historical land uses and their concentrations in receiving streams can remain high for many years because of the slow movement of ground water.

The hydrologic conditions under which constituents enter a stream are important when assessing the source of nutrient loads to the bay. Concentrations of constituents in streams of the bay watershed are a summation of the nonpoint source contributions from constant ground water discharge and intermittent storm runoff sources. Ground water contributions to a stream are relatively constant, varying only slightly with season. Concentrations of constituents carried to a stream by ground water can be quantified in samples collected during base flow conditions. Storm runoff, composed of overland runoff and interflow, contributes to a stream intermittently, depending on storm intensity and frequency, and only during high flows. Storm runoff dilutes the ground water contributions to a stream. Constituents intermittently carried to a stream by storm runoff along with the constant contributions from ground water can be quantified in samples collected during storm flow events.

The magnitude or steepness of the regression slope of constituent load to stream flow can be used to indicate the relative contributions of ground water and storm runoff at a river location (Hunchak-Kariouk et al., 1999; Price and Schaefer, 1995). The steeper the slope, the greater the contribution from nonpoint sources during increased stream flow. If the contributions to instream load are predominantly from ground water, instream load will remain relatively constant with increasing stream flow and the regression slope of load to stream flow will be approximately zero. If, however, storm runoff contributes a disproportional amount to instream load, instream load will increase with increasing stream flow and the regression slope of load to stream flow will be greater than zero. Comparison of the relative magnitude of the load slopes between stations and for different constituents at a station can indicate the relative importance of storm runoff and ground water to instream loads at a station for each constituent.

The amount of total phosphorus (TP) entering the bay via river discharge is small compared to the amount of nitrogen, measured as nitrite plus nitrate and total ammonia plus organic nitrogen. Total phosphorus load appears to be related to the type of land cover - basins with greater urban land and agricultural plus grasslands cover have somewhat larger concentrations, yields, and loads of TP. Yields calculated for monitored basins and estimated for unmonitored areas are larger for basins classified as Landscape II (> 10% development) than as Landscape I (< 10% development). About one-third of the load comes from the unmonitored coastal areas (mostly Landscape II) of the basin. Yields are largest at the most upstream and downstream sites on the Toms River, the downstream site on Wrangel Brook, and the site on Oyster Creek.

Nonpoint storm runoff does not seem to be a significant source of TP to the bay because concentrations are small, especially in areas of slight urban land cover, and either the same or larger during low flow than high flow in areas of moderate to high urban land cover. Using the values presented by Hunchak-Kariouk et al. (1999) for relations between water quality and stream flow at USGS stations in the Atlantic Coastal Plain, the slope of TP load to stream flow was higher at McDonalds Branch than at Toms River near Toms River, which was slightly higher than at Great Egg Harbor near Sicklerville. Storm runoff is a greater contributor to instream load at McDonalds Branch than at Toms River or Great Egg Harbor River. The Great Egg Harbor River Basin is more developed than the basins of Toms River and McDonalds Branch and the concentrations were slightly larger at Great Egg Harbor than at Toms River and McDonalds Branch.

Total nitrogen (TN) is a measure of several nitrogen species that can be dissolved or associate with particles. The concentration and loading of TN in a stream and the relative importance of groundwater and storm runoff contributions depend on which nitrogen species is predominant. About half the TN measured in surface water of the Bay watershed is NO2 + NO3; the other half is organic nitrogen because total ammonia concentrations are small in most basins. The amount of NO2 + NO3 in a stream is strongly related to the intensity and type of land cover in the contributing area basins. Greater urban land and agricultural plus grasslands cover have much larger concentrations, yields, and loads of NO2 + NO3 than less developed basins. Yields calculated for monitored basins and estimated for unmonitored areas are in general an order of magnitude larger for basins classified as Landscape II (greater than 10% development) than as Landscape I (less than 10% development). Yields calculated at all sites on Wrangel Brook were especially large. The most downstream area of the Wrangel Brook basin had nd estimated NO2 + NO3 yield of greater than 1,900 (lb/mi²/yr). Yields from direct groundwater discharge from nearby areas with no major streams are also estimated to be large.

Ground water appears to be a significant source of NO2 + NO3 to the bay because concentrations are larger during low flow than high flow. Concentrations during low flow and high flow are larger in areas with moderate to high urban land cover than in areas with slight development.

The remaining half of the TN in surface water is organic nitrogen; NH4 concentrations were small except at the site on Mill Creek where it was a very large component (more than half of the TN). Nonpoint storm runoff is most likely an important source of TAON to the Bay because concentrations are larger during low flow than high flow, especially in areas of moderate to high urban land cover. TAON also seems to be related to land use; concentrations are larger in areas with moderate to high amounts of urban land cover and wetlands. The yield of NH4 and TAON from the Mill Creek was especially large.

Nonpoint storm runoff does not seem to be a significant source of NH4 to the Bay, except in areas of greater urban land cover. Median concentrations during all flows were low (0.08 mg/l or less) at all sites except 4 of the 10 most developed sites. Of the sites with median concentrations greater than 0.08 mg/l, the median concentrations during low flow were the same or greater than during high flow at the two most-downstream stations on the Toms River and the Mill Creek station.

VIII.  TOTAL NITROGEN AND NITRITE PLUS NITRATE LOADS TO THE ESTUARY

Total loads to the BBLEH can be computed as the sum of the surface-runoff loads and the ground water discharge. For NO2 + NO3, the total annual load is 560,000 lb/yr 252,000 kg/yr) -- 360,000 lb/yr (162,000 kg/yr) from surface runoff and 200,000 lb/yr (90,000 kg/yr) from ground water discharge (rounded values).

The overall yield of NO2 + NO3 from surface water and ground water discharge to the BBLEH is estimated to be 1,000 lb/mi²/yr and the yield from surfacewater discharge was 900 lb/mi²/yr. The yield of TN from surface water discharge was 2,000 lb/mi²/yr. About half of the nitrogen load discharged to the BBLEH from rivers is organic nitrogen. This load of organic nitrogen to the BBLEH may be significant.

Yield values of total nitrogen and nitrite plus nitrate from the BBLEH watershed are comparable to those estimated for other Coastal Plain basins. For the BBLEH watershed, nitrite plus nitrate and nitrate yields can be assumed equivalent because nitrite concentrations in surface and ground water were small. Yields of nitrate to the Chesapeake Bay were estimated to be 2,400 lb/mi²/yr from total flow (total stream flow) and 1,700 lb/mi²/yr from base flow; total flow, total nitrogen yields were 4,460 lb/mi²/yr (Bachman et al., 1998). Yield values of total nitrogen were estimated to be 2,900, 1,600, and 4,600 lb/mi²/yr from three basins in the Patuxent River basin in the Maryland Coastal Plain (Preston, 1996). Base flow yield values of total nitrogen were estimate to be 3,800 and 900 lb/mi²/yr, depending on the method used (Bachman and Phillips, 1996).

IX.  CHANGES TO DESIGNATED HUMAN USES IN THE WATERSHED

Historical increases in population and water demand in the Barnegat Bay watershed have resulted in or contributed to substantial water supply problems, including saltwater intrusion, stream depletion, and concerns about the quality of drinking water. Restrictions on the use of confined aquifers in order to reduce the threat of saltwater intrusion have resulted in an increased reliance on surficial aquifers and surface water to meet increasing water demands. The implementation of these restrictions has raised concern about environmental impacts that can result from excessive stream depletion and the vulnerability of water supplies to contamination.

Prior to the development of regional water supplies, water pressure levels in all confined aquifers in the New Jersey Coastal Plain were above sea level. Withdrawals from confined aquifer supply wells located within and outside the project area have produced cones of depression that extend over large parts of the watershed area. Withdrawals from major confined aquifers in the northeastern coastal plain have also resulted in the migration of saltwater into two productive aquifers in northern Monmouth County and in Middlesex County (USGS, 1989).

Withdrawals from confined aquifers in the northeastern coastal plain region have resulted in deep cones of depression in areas where relatively low rates of groundwater flow from adjacent areas and overlying aquifers are available to replace the water withdrawn. By 1983, confined aquifer withdrawals in response to high water demand had caused groundwater pressure levels to decline to more than 200 ft (61 m) below sea level in some areas (USGS, 1995a). If withdrawals had been left unrestricted, groundwater levels would have declined even further.

Saltwater intrusion caused by excessive confined aquifer withdrawals has resulted in elevated chloride concentrations in supply wells in the upper and middle aquifers of the Potomac-Raritan-Magothy Aquifer system to the north of Barnegat Bay. Specific areas where the upper aquifer is affected by saltwater include South Amboy, Keyport, Union Beach, and Keansburg. Specific areas where the middle aquifer is affected by saltwater are Sayreville, South Amboy, Keyport, and Union Beach (USGS, 1996).

In response to a similar saltwater problem in the west-central part of the New Jersey Coastal Plain, the New Jersey Department of Environmental Protection designated Water Supply Critical Areas 1 and 2 (Figure 13), and mandated reductions in confined aquifer withdrawals in these areas. Since reductions in Critical Area 1 were initiated in 1990, groundwater levels in the northeastern coastal plain have substantially recovered, but the threat of saltwater intrusion remains.

Although saltwater intrusion into the major confined aquifers is not known to be a problem for supply wells located in the Barnegat Bay watershed, supply wells located in several areas within the watershed area appear to be potentially threatened by saltwater intrusion in the long term (O. Zapecza, U.S. Geological Survey, written communication, July 6, 1994).

These areas include:

•Long Beach Island (Atlantic City 800-Foot Sand Aquifer);

•Barnegat Light, Seaside Heights, and Seaside Park (Piney Point Aquifer);

•Point Pleasant, Lavallette, and other Monmouth County locations (Englishtown Aquifer system);

•Point Pleasant, Chadwick, and Lavallette (upper aquifer of the Potomac-Raritan-Magothy Aquifer system); and

•Lavallette, Toms River, and other locations in northern Ocean County (middle aquifer of the Potomac-Raritan-Magothy Aquifer system).

Because the major confined aquifers are overlain by thick layers of clay and silt, they are generally less vulnerable than unconfined aquifers to contamination resulting from human activities at the land surface.

Many municipalities, public purveyors, and private well owners use the unconfined Kirkwood-Cohansey aquifer system. Because of the potential threat of saltwater intrusion in the confined aquifers of the region, it is anticipated that requests for additional withdrawals from the Kirkwood-Cohansey Aquifer system for public supply and irrigation uses will increase. Consumptive withdrawals from the unconfined aquifer system can reduce stream flow (Figure 14), resulting in habitat loss and a reduced capacity of streams to dilute contaminant loads. The unconfined aquifer system is also much more vulerable to contamination due to human activities at the land surface, and numerous instances of shallow groundwater contamination in the Barnegat Bay watershed area have been documented (USGS, 1984, 1997a, b, 1998; Ocean County Health Department, written communication, 1998).

Historic groundwater withdrawals have caused the shifting of flow patterns in parts of the Kirkwood-Cohansey Aquifer system, reducing groundwater discharge to streams and wetlands of significant ecological value. A recent investigation of trends in stream discharge over time concluded that stream flow measured at gaging stations on the North Branch of the Metedeconk River and Toms River decreased periodically between 1970 and 1989 (USGS, 1995b). However, the cause(s) of these declines has not yet been determined. In addition, there has not been a determination of whether ecological resources or shellfish were affected by the reduced flow. Nevertheless, the New Jersey Department of Environmental Protection is presently evaluating actions that should be taken to minimize possible impacts.

Recent hydrogeologic investigations indicate that withdrawals from the Kirkwood-Cohansey have lowered groundwater levels and have caused average reductions in stream flow of up to 11% (USGS, 1997a). Stream flow in some streams during natural low flow periods could decrease by ~26%, if water allocations are fully utilized. Kettle Creek, in particular, could be most affected. Withdrawals have caused the average groundwater level to decline by up to 20 ft (6.1 m) near pumping centers. At full allocation, it is estimated that water levels can decline by an additional 20 ft (6.1 m). These impacts can potentially damage riverine ecosystems and reduce the yields of the surface water supplies of the Metedeconk River (and the Toms River if it is used for water supply purposes in the future).

There are two areas in the region where saltwater intrusion has affected wells withdrawing water from the Kirkwood-Cohansey Aquifer system. Saltwater has adversely affected public supply wells located in Seaside Heights and Point Pleasant Beach, where observed chloride concentrations in groundwater have exceeded the 250 mg/l secondary drinking water standard (Figure 15). There are numerous other public and private wells that are located near brackish water along the coast. Consideration should be given to additional monitoring, where applicable, especially if higher allocations are proposed.

Domestic wells in Berkeley, Dover, Eagleswood, Lakewood, and Stafford produce water with high levels of sodium (Bergman and Associates, 1990). It is unknown if this phenomenon is caused by saltwater intrusion, road salting, or septic systems.

Present and future withdrawals from the region's confined aquifers can also shift groundwater flow patterns and result in stream flow depletion. However, these effects have not been quantified; therefore, the cumulative effects of all pumpage are not well understood.

Ocean County has the largest number of domestic wells of any county in New Jersey. It is thought that large numbers of these wells are used for lawn irrigation. When there are large numbers of domestic wells that are used for this purpose in close proximity to each other, the effects can be similar to a large public well that serves homes and businesses that are in a sewer service area. Much of irrigation water is subject to evaportranspiration and is not "recycled" into the aquifer system. It was estimated that domestic wells in Ocean County used 10.7 MGD in 1990 (Bergman and Associates, 1990). Most of these wells are in the Kirkwood-Cohansey Aquifer system.

The intrinsic link between anthropogenic disturbance and changes in aquatic community structure has been consistently documented over the past decade. Land-use alterations resulting in an increase in impervious surfaces, runoff, suspended sediment, and nutrient loading directly affect the hydrology, geomorphology, and water quality of the Barnegat Bay watershed, and alter the aquatic communities that inhabit this system. Because this link exists, it is possible to assess the severity of anthropogenic disturbances by comparing aquatic communities (for example, aquatic macroinvertebrates) found at specific sampling sites to that expected at reference or minimally impaired sites and by relating measured environmental conditions to aquatic communities that have been modified.

Many state and federal government agencies have initiated biomonitoring programs to assess the effects of anthropogenic disturbances on aquatic communities (e.g., Loeb and Spacie, 1994; Davis and Simon, 1995). In 1992, the New Jersey Department of Environmental Protection (NJDEP) initiated a watershed-based biomonitoring program known as the Ambient Biomonitoring Network (AMNET) to evaluate the biological integrity of aquatic invertebrate communities in New Jersey streams including tributaries of the Barnegat Bay-Little Egg Harbor estuary. This statewide network incorporates regional reference sites, which are minimally disturbed areas organized by selected chemical, physical, and biological characteristics (New Jersey Department of Environmental Protection, 1994; Reynoldson et al., 1997). The goal of the program is to monitor the condition of benthic macroinvertebrate communities every five years. This frequency is considered to be realistic for evaluating environmental changes. Sampling locations were chosen for monitoring mainstem locations above major tributary confluences, assessing the effects of lakes, investigating known sources of contamination, and evaluating the effects of significant natural features such as wetlands, preserves, and wildlife management areas. Concurrently, the U.S. Geological Survey in cooperation with the NJDEP operates a statewide surface water quality monitoring network that provides the basis for the state water quality inventory report to the U.S. Environmental Protection Agency and Congress mandated by the Section 305(b) of the Clean Water Act. Together, these programs maximize the integration of water quality and biological information and provide a solid foundation for statewide water quality planning and management decisions involving surface water quality standards and biocriteria in New Jersey.

The bioassessment method chosen by the NJDEP for evaluating aquatic macroinvertebrate communities was modified from the U.S. Environmental Protection Agency's Rapid Bioassessment Protocol II (Plafkin et al., 1989). Rapid bioassessments are based on a multimetric approach which uses an array of individual measures (e.g., community, population, structural, and functional) to summarize diverse biological information into a single measure of condition. Individual measures of biological condition (referred to as metrics) for each sampling site are compared to a statewide reference database (New Jersey Department of Environmental Protection, 1994). On the basis of this comparison, a total community "impairment score" is established. Benthic macroinvertebrates are ideally suited for this type of approach because they are sensitive to minor changes in water quality and therefore are useful as indicators of a wide range of environmental disturbances.

Five metrics comprise the NJDEP's rapid bioassessment protocol (total taxa richness, total EPT richness, percent dominance, modified family biotic index, and percent EPT) (Table 18). The primary requirements for inclusion of metrics in the New Jersey RBP are low variability, the demonstrated ability to discriminate non-impaired from impaired conditions, and statistical independence (non-redundancy). A brief description of metrics used in the New Jersey RBP follows. The total taxa richness measures the total number of families identified in the sample. A reduction in richness may indicate environmental stress related to organic enrichment, toxics, and sedimentation. The EPT richness index measures the total number of ephemeropteran (mayfly), plecopteran (stonefly), and trichopteran (caddisfly) families in a sample. These organisms are considered to be highly sensitive to disturbance, and typically, the number of families vary inversely with the magnitude of environmental disturbance. The percent dominance assesses the relative balance within a macroinvertebrate community. Healthy macroinvertebrate communities are often characterized by an equitable faunal assemblage. Degraded streams are commonly dominated by a few highly tolerant taxa or taxa groups. The modified family biotic index (FBI) was developed to evaluate the relative tolerance of benthic macroinvertebrates to organic enrichment (Hilsenhoff, 1988). This index is based on a gradient or continuum of tolerance that ranges from 0 (sensitive) to 10 (tolerant) and typically increases as water quality decreases (Plafkin et al., 1989). The percent EPT provides a measure of the percent abundance of three sensitive aquatic insect families. A high percentage of Ephemeroptera, Plecoptera, and Trichoptera in a sample is associated with good water quality. Percent abundance of these groups often decreases with only minimal increases in environmental degradation.

In the Barnegat Bay watershed, benthic macroinvertebrate samples were collected from August to December 1994. To maintain consistency among sampling sites, a multihabitat sampling approach was used. Benthic macroinvertebrates were sampled with a Surber sampler (box-type sampler) or a rectangular kick net. A grab sampler was used to collect benthic macroinvertebrates located in depositional areas. In addition, a separate sample of coarse particulate organic matter (primarily decomposing leaf litter) was collected by hand. All material from the samples was combined, and a family-level 100-organism subsample was randomly generated from the entire sample. Three condition categories (non-impaired, moderately impaired, and severely impaired) were established to reflect the type and level of macroinvertebrate community impairment in New Jersey streams (see Table 18 for scoring criteria and definition of the three impairment ratings).

Sixty-one sites were sampled within the Barnegat Bay watershed as part of the AMNET program. These sites range in condition from non-impaired to severely impaired (Table 19, Figure 16). Of the 61 sites sampled, 57% were classified as non-impaired, 36% as moderately impaired, and 7% as severely impaired (Figure 17). Three of the four sites classified as severely impaired (AN0525, AN0532, and AN0510; Table 19, Figure 17) are proximally located to hazardous waste sites within the Barnegat Bay watershed. The percent EPT ranged from zero at AN0525 and AN0532 to 91% at AN0548. The modified FBI, which increases with increasing levels of organic contamination, ranged from 2.02 at AN0548 to 8.91 at AN0525. The most diverse site was AN0507 with 57 different taxa and the least diverse sites were AN0525 and AN0532 with four taxa (Table 16).

Twenty of the 61 benthic macroinvertebrate monitoring sites in the Barnegat Bay watershed were directly linked to cooperative water quality monitoring stations. Detrended Correspondence Analysis (DCA, an indirect ordination technique) was used to assess potential land-use and water quality gradients in the watershed (Figure 18). DCA is an ordination technique favored by many practical field ecologists for non-linear data analysis of a variety of organisms including plant communities (Chang and Gauch, 1986), benthic organisms (Wright et al., 1984; Leland et al., 1986; Tate and Heiny; 1995), terrestrial vertebrates (Ben-Shahar and Skinner, 1988), and fishes (Edds, 1993; Galacatos et al., 1996; Zampella and Bunnell, 1998). It is considered to be one of the most powerful multivariate tools available for assessing patterns in communities composed of species that vary along compositional gradients (Peet et al., 1988). The underlying assumption of this approach is that species respond unimodally to environmental constraints. More specifically, species abundances tend to increase with an increase in some environmental parameter, reach a maximum at some environmental optimum, and then decrease in abundance at increasing levels of the environmental parameter. Therefore, if any aspect of the environment is greater or lesser than this optimum, the species will perform more poorly (i.e., it will have reduced abundance). This technique effectively summarizes community variation relative to underlying gradients and is scaled in units of average standard deviation (SD) or species turnover (Gauch, 1982). A species appears, rises to its mode, and disappears over a length of approximately 4 SD's, which is equal to one species turnover (beta diversity). This is a measure of how sites differ from each other or, alternatively, it is a measure of the length of an ecological gradient or ordination axis in terms of species composition. An axis or gradient with high species turnover will have completely different species compositions at opposite ends of the ordination diagram. Thus, ordination by DCA arranges sites with similar taxonomic composition to cluster more closely together and sites with dissimilar taxonomic composition to cluster farther apart (Gauch, 1982).

Site scores produced from this ordination procedure can be related to environmental variables. The site scores represent a coordinate along an ordination axis specifying the x-y location of the sample in ordination space. Ideally, site scores represent the position of communities along an important environmental gradient. Each of the DCA axes comprises part of the basic structure of the Cartesian coordinate system, and the axes are usually portrayed as being at right angles to each other (orthogonal; Figure 19). The axes for DCA are obtained by using a statistical procedure known as reciprocal averaging. This technique involves matrix algebra and simultaneously arranges species and sites by calculating the weighted average or centroid of the species and site scores in ordination space. This is an iterative approach which continues for each axis until the scores stabilize (converge to a unique solution). Convergence is relatively rapid; however, this is dependent upon the initial size of the data matrix (Digby and Kempton, 1995). Thus, the scores themselves form a gradient because they are ordered along each axis and the importance of an axis is commonly represented by its eigenvalue. The eigenvalue is actually equal to the maximized dispersion of the species scores along the ordination axis and is thus a measure of the relative strength of the ordination axis (Jongman et al., 1995). Eigenvalues always range between 0 and 1. The first axis of an ordination will have the largest eigenvalue, the second axis the second largest eigenvalue, and so on (Jongman et al., 1995).

AMNET sampling sites were linked with corresponding water quality data and used to assess potential anthropogenic changes in the Barnegat Bay watershed by subsequently comparing the site scores of DCA axes 1 through 4 to environmental variables using Spearman rank correlations (Table 20). Only the first 4 axes of the ordination generally are interpreted because, in practice, they represent or account for most of the variation in the species data. One goal is to determine whether site scores are related to anthropogenic changes in the environment. By using this approach, it is possible to determine indirectly which environmental variables contribute most to the separation of sites along a particular axis (Table 20), that is, which environmental variables exert the greatest influence in structuring the aquatic communities inhabiting tributaries of the Barnegat Bay.

As described earlier, the relative magnitude of the eigenvalues for each DCA axis describes the relative importance or strength of the environmental gradient and the length of the gradient is a measure of how unimodal the species responses are along an ordination axis. A gradient can be defined as a spatially varying aspect of the environment, which is expected to be related to species composition. It has been suggested that a gradient length of 4 SD's represents a full species turnover, and that sites at the opposite ends of the axis would have few species in common (Jongman et al., 1995). The gradient length for the ordination of sites in the Barnegat Bay watershed was 6.01, representing a strong unimodal response (Figure 18) and more than one species turnover. This suggests a low similarity between sites located at opposing ends of the axes in the ordination diagram (Figure 19).

Twelve environmental variables were found to be significantly correlated with the four DCA axes (Table 20). The first axis of the ordination diagram (x-axis, Figure 20) represents a separation of sites based on the number of taxa present at each site (a diversity gradient). High diversity suggests that niche space, habitat, and food sources are adequate to support the survival and propagation of the endemic fauna. Those sites located on the right side of the ordination diagram have the highest species diversity and those on the left, the lowest.

The second ordination axis, which runs from the top to bottom of the ordination diagram (y-axis, Figure 19), reflects a strong water quality impairment gradient and separated sites based primarily on community metric criteria. Three variables, the New Jersey Impairment Score (NJIS, which is the numerical sum of all metrics shown in Table 18), FBI, and percent EPT were significantly correlated with the second axis (Table 20). In general, sites located at the top of the second axis are least impaired and those at the bottom are most impaired. The presence of EPT taxa (mayflies, stoneflies, and caddisflies) is strongly associated with good water quality conditions, and it is well established that these three families of aquatic insects typically decrease with only minimal increases in environmental degradation. The premise underlying composition based metrics such as %EPT is that a healthy and stable macroinvertebrate assemblage will be relatively consistent in its proportional representation, though individual abundances may vary. Thus, a high percentage of EPT taxa in a sample would represent sites with good water quality; for example, those sites at the top of the ordination diagram (AN0524, AN0535, AN0523, AN0519, and AN0547) typically exhibit unimpaired conditions and better water quality than those at the bottom of the ordination diagram (AN0510, AN0532, AN0538, AN0509, AN0518). The modified FBI was established to assess the relative tolerance of benthic macroinvertebrates to organic enrichment. This index is based on the premise that some organisms are more tolerant of a variety of stressors than others. Some organisms are considered to be sensitive to contamination, and others are indicative of a highly stressed system. For example, many Ephemeroptera (mayflies) are highly intolerant organisms, and their presence may be indicative of a site that has little organic enrichment. However, the dominance of members of the caddisfly family Hydropsychidae is indicative of a stressed system. AN0548 has the highest percentage of ephemeropteran taxa (91.1%, Table 19) and is located at the top of ordination diagram (Figure 19). AN0532 and AN0510, however, have no ephemeropteran taxa and are nearest the bottom. The NJIS was designed to incorporate the characteristics of benthic macroinvertebrate assemblages that are best able to distinguish water quality impairment. The finding that the second DCA axis is significantly correlated with the NJIS as well as %EPT and FBI (Table 3) further substantiates the strong water quality gradient that exists within the Barnegat Bay watershed.

The third axis (not shown in the ordination diagram) represents a wetland gradient that reflects also a change in pH and flow characteristics. Sites with the highest percentage of wetlands in the watershed appear to have the lowest pH. These factors were found to be strongly influenced by latitude which was also significantly related to the third axis (R=0.51, p=0.02). This suggests, in general, that sites nearer to the eastern side of the basin and more proximally located to urban areas and the Garden State Parkway have a lower percentage of wetlands and higher pH. It is well documented that changes in pH affect the reproduction and survival of endemic aquatic organisms. Thus, as urban corridors continue to increase in the Barnegat Bay watershed, changes such as the reduction in the amount of wetlands may produce measurable negative consequences for stream water quality and the aquatic communities that inhabit these systems.

The fourth axis scores were found to be strongly related to median specific conductance (MD_SC), flow, number of individuals (NOINDIV = abundance), the ratio of EPT to the family Chironomidae (EPT/C), latitude, and percent forest (Table 20). This was the weakest of the gradients assessed; however, the eigenvalue for the fourth axis was fairly strong and still accounted for a significant amount of the variance. MD_SC, flow and NOINDIV were the three most strongly correlated environmental parameters with axis four. MD_SC and NOINDIV were negatively correlated, however, flow was positively correlated with the fourth axis. These relations suggest that as flow increases, specific conductance and abundance of aquatic macroinvertebrates in tributaries of Barnegat Bay decrease. Runoff from urban development has been directly linked to changes in specific conductance and stream flow characteristics. Although these factors can vary seasonally, the impact is measurable and can directly affect the abundance of benthic organisms in streams of the Barnegat Bay.

The biological impairment associated with physically degraded sites is clearly related to a change in landuse in the Barnegat Bay watershed. Changes in land-use practices appear to have impaired the physical and biological integrity of many of these streams by decreasing the quality and quantity of specific habitats in the basin. As land is further developed for residential and commercial uses, important basin characteristics (e.g., the amount of area classified as wetlands) are significantly reduced. In addition, development alters flow regimes by straightening, deepening, and diverting natural channels. These changes in conjunction with the reduction in wetland habitat likely explains the strong correlation between the third axis and flow characteristics found for streams in the Barnegat Bay (R=0.46, p=0.04). Rain falling on vegetated surfaces can be absorbed more rapidly than on paved surfaces, and it is well established that wetlands mitigate flood peaks. Rain falling on impervious surfaces, however, is not absorbed, and thus can sharply increase runoff peaks, which can negatively affect the aquatic communities that inhabit these streams.

Many of the factors linked to the degradation of benthic communities are becoming better understood. Factors such as the amount of impervious area (areas in a watershed covered by asphalt, concrete, buildings, and other compacted surfaces), urban runoff, increased suspended-sediment loads, habitat degradation, sewage effluent, and greater urbanization all have been shown to affect the hydrology, geomorphology, and water quality of stream systems. In the Barnegat Bay watershed, aquatic macroinvertebrate communities appear to be driven by the level of organic enrichment and the percentage of wetlands present in a basin. This is indicated by the strong water quality impairment gradient found along the second ordination axis and the significant correlations with wetlands, stream flow, and pH.

Barnegat Bay, Manahawkin Bay, and Little Egg Harbor comprise a unique lagoon-type estuarine system (Figure 20). An array of environmentally sensitive habitats exists in these shallow bays, such as waterfowl nesting grounds, salt marshes, submerged aquatic vegetation, finfish nursery areas, and shellfish beds. Biotic communities are replete with rich assemblages of planktonic, nektonic, and benthic organisms that are highly responsive to variations in natural and anthropogenic factors. Some of these organisms (e.g., winter flounder, Pseudopleuronectes americanus; hard clams, Mercenaria mercenaria; blue crabs, Callinectes sapidus) are also of recreational and commercial importance.

Human activities both in watershed areas and on open bay waters have affected sensitive habitats and living resources of the estuary. Many anthropogenic impacts in Barnegat Bay are directly coupled to development in the bay watershed and to activities associated with human habitation, notably construction, land maintenance, agriculture, and vehicle use. Watershed modifications that have particularly affected estuarine water quality include deforestation and infrastructure development together with landscape partitioning and paving. Along the estuarine perimeter, marsh filling and bulkheading, diking and lagoon construction, and dredging and dredge material disposal have disrupted natural habitats and altered biotic communities. Human activities in the watershed significantly alter the timing, magnitude, and nature of material inputs to the estuary.

The Barnegat Bay-Little Egg Harbor estuary is highly susceptible to pollution because of its limited dilution capacity and flushing rate. Major groups of chemical contaminants occurring in sediments, biota, or estuarine waters are halogenated hydrocarbons, polycyclic aromatic hydrocarbons, trace metals, and radioactive substances. These contaminants, which are generally poorly characterized, may derive from both point sources (e.g., industrial wastewater discharges, marinas, and dredged materials) and nonpoint sources (e.g., suburban and construction-site runoff, faulty septic systems, groundwater pollutant transfer, boats, and atmospheric deposition). Their effects on organisms in the estuary have not been addressed. Operation of the Oyster Creek Nuclear Generating Station is not only responsible for biocidal and radioactive contaminant releases into the estuary but also for increased mortality of estuarine organisms due to thermal discharges, as well as impingement and entrainment. Nearly all previous investigations of anthropogenic impacts on biotic communities in the estuary have focused on power plant activities.

Among the highest priority problems encountered in this system are those associated with nutrient and organic loadings. Barnegat Bay is classified as a moderately eutrophic system (Seitzinger and Pilling, 1992, 1993). Nutrient enrichment leads to elevated phytoplankton biomass and production in the summer. Seasonal phytoplankton blooms occur in the estuary, with winter blooms dominated by diatoms (e.g., Thalassiosira nordenskiolkii and Detonula confervacea) and summer blooms dominated by a nonmotile chlorophycean (green) alga (i.e., Nannochloris atomus). Brown tides comprised largely of a coccoid picoplankter, Aureococcus anophagefferens, have been documented in the southern part of Little Egg Harbor, most recently during the summers of 1995, 1997, and 1999 (Mahoney et al., 1997; Olsen, New Jersey Department of Environmental Protection, personal communication, 1999). The accumulation and bacterial decay of phytoplankton remains on the estuarine floor during bloom events are a a concern because they may promote hypoxia of estuarine bottom waters that can be detrimental to biotic communities.

Additional impacts are attributable to pathogens (e.g., coliform bacteria) from animal waste washed into the bay along roadways and adjacent lands, from bird droppings, as well as from human waste from faulty septic systems and boats. Elevated coliform bacteria levels periodically result in beach closings and the closure of shellfish beds to harvesting. Apart from these impacts, increased turbidity and siltation levels, particularly in tributaries of the estuary, result from construction, dredging, and bank stabilization projects. Finally, the estuary is frequently littered with floatable debris ascribed to stormwater runoff, beachgoers, recreational boaters, and various illegally dumped sources.

The Barnegat Bay-Little Egg Harbor estuary is plagued by many of the same e